Speed
No flicker
The version is chosen on the server before the page reaches the visitor. A visitor never sees A for a moment before B.
Product A/B tests
Show two versions and keep the one that brings more
Make a second version of a page in the same builder. Husly shows A to some visitors and B to others, counts the enquiries and tells you whether the gap is real. One click puts the winner on your page.
Version B
A new version in the same builder
B starts as a copy of your page. Change the heading, the picture or the button as on any other page. One test can have up to four versions, and you decide what share of the visitors sees each one.
Version B
B is a copy of your page. Change it in the same builder as any other page, up to four versions, A to D.
Change one thing
Another heading, another picture or another button. Your live page does not change until you pick the winner.
The goal
You decide what counts
A click on a button, a visit to a thank-you page or a form sent. Or something anywhere on your site: a chosen form, a booking, a chat or feedback. The server counts those itself, so nobody can fake them.
Published pages only
A draft page gets no visitors, so a test runs only on a published page.
Anywhere on your site
A form, a booking, a chat or feedback counts even when the visitor does it later that day on another page.
That very form
Choose which form is the goal, for example the sign-up for a free trial class.
The results
Is the gap real or luck
Each version is compared with A. Husly says better or worse only when there is enough data and the gap is at least 95% sure. Until then you see needs more data, so you do not crown a winner too early.
Better, 95% sure
A verdict shows only when both versions have at least 100 visitors and 5 conversions.
Worse is an answer too
A version that is clearly worse is red, so a poor change never reaches your page by mistake.
Day by day
The chart shows each version's conversion rate from the start of the test.
The schedule
The test stops by itself
Set an end date, or let the test stop as soon as the result is clear. You get an email and a notice about a clear result, also when the test keeps running.
An end date
The test stops by itself at the end of that day.
Stop when the result is clear
The test stops as soon as a version is clearly better or worse than A. You get an email and a notice.
The winner
The winner goes live in one click
When the result is clear, pick the winner. Husly asks first, because B replaces your live page, and says exactly what will happen.
A plain question
Before B replaces your live page, Husly says exactly what will happen.
One click
The winner goes live at once and the test ends.
The parts that work in the background
A good test is an honest test: the right visitors, the right numbers and a page that does not flicker.
Search engines
Google sees the original
Robots and search engines always see version A and never count. The title, the description and the canonical address always come from the original.
Privacy
Only with consent
Only a visitor who allows analytics cookies takes part. The others see version A. Starting a test switches the cookie banner on, and Husly tells you so.
Connections
The result reaches other tools too
A webhook sends an event when a result is clear or a test ends.
- Zapier and Make
- or your own system
Questions about A/B tests
How much do A/B tests cost?
A/B tests are in the Standard plan and every plan above it, from €14 a month. There is no limit on tests; one test at a time can run on a page.
How long does a test take?
As long as it takes to get enough data: at least 100 visitors and 5 conversions on both versions, and a gap that is at least 95% sure. On a small site that takes weeks. Set an end date, or let the test stop when the result is clear.
Does a test hurt my place in Google?
No. Search engines always see the original, and the title, the description and the canonical address stay the same. The address does not change; every version lives at the same address.
What can I test?
A whole page: the heading, the pictures, the buttons, the order of the blocks or a whole other design. Up to four versions at once, each one in the same builder.
Do I need to ask for cookie consent?
Husly does it for you. Only visitors who allow analytics cookies take part. If the site has no cookie banner yet, starting a test switches it on.
What happens after the test?
Pick the winner. When it is B, C or D, it replaces your live page at once. When A stays, nothing changes. Husly asks first either way.
Start your first test
Pick a page, make a version B and see which one brings more enquiries. A/B tests are in the Standard plan.